## BE Analytics: The recipe for winning UFC Fight of the Night bonuses

Bloody Elbow’s analytics expert shares the formula for winning UFC Fight of the Night and some potential dangers of relying on simple statistics.

Take a look at these two fights. Each took place during the same event and both went to a three-round decision.

Which looks like the Fight of the Night (FOTN) to you? You'd probably pick Fight A due to the large difference in strikes thrown and landed as well as more time spent standing and banging, right? In fact, the award winner was Fight B. So how does the UFC decide which fight wins FOTN? What are the key factors?

I decided to do this piece after being dissatisfied with the way FOTN had been analyzed previously. It looked something like this:

Bout order "1" is the main event, "2" is the co-main event, and so on. The main card is typically five bouts which would be spots 1-5 in the table. The prelims would typically be spots 6 and above. What did you learn from this chart? It definitely looks like more awards go to main card fights, and the most clearly go to the main event, right? But that ain't science, folks. That's just screwing around with simple averages, which can sometimes be misleading. It actually turns out that while there is a difference between the main card and the prelims, there's no statistical difference between the main event and the rest of the main card.

Simple analyses tell us what happened but don't tell us anything about why. Are fights that are likely to be more entertaining put on the main card? Are title fights - which are mostly in the main event spot - more likely to win FOTN since people are excited for them? If all fights were equal then simple is fine, but all fights definitely aren't equal.

Today we're crescent-kicking simplicity and examining the why. I'm using a model that economists employ quite frequently which reveals the factors that help, hurt, and do nothing for your chances of winning FOTN. Here's what I find after applying it to every UFC event with a FOTN award through the end of 2013 - almost 2,000 fights in all.

Close Fight

The closeness of a fight can be statistically measured. A fight that's back-and-forth close or tight the entire way has a much better chance of winning FOTN than a one-sided affair, even if it's an incredibly dominant performance by one fighter. Close fights tend to be exciting fights and a very close fight is worth 100 points towards FOTN. As a fight becomes more one-sided, the points earned decline towards zero.

Idle Time

Fights with a lot of idle time aren't very entertaining, so every minute with no activity lowers your odds of winning FOTN. But it matters where the idle time takes place. Each idle minute at distance, in the clinch, and on the ground costs the fighters 8, 22, and 10 points, respectively. The most obvious takeaway is that while all idle time hurts, the clinch is your mortal enemy if you're looking to win FOTN. Most clinch time is spent pressed against the cage which isn't terribly exciting. So if you like money, "Un-clinch yourself!" It's better to take a nap on the other guy than to press him against the cage.

Activity

So you wanna be a FOTN'er? The striking recipe is simple: hunt for the head with punches, elbows, knees and kicks and don't worry too much about connecting. If you do connect, jabs and power shots to the head are worth one and two points, respectively, while knockdowns are worth 38. Recall last week's article about how certain missed strikes score points with the judges towards winning a round? Well they can also help win FOTN as missing a power shot to the head is worth one point.

The only other way to earn striking points is by missing a power shot to the legs (4 points). I know, four points seems high compared to the other strikes. What I think is going on here is that missed power kicks are rare, can be more showy (fighter goes spinning) and are more likely to get Oooh's and Ahhh's from the crowd than missed power punches. The vast majority of missed power leg strikes are kicks so the full effect is revealed, whereas very few head shot misses are kicks so the effect is drown out by the other more standard strikes. If you have a different idea, share it in the comments. BTW, saying that I'm an idiot is a perfectly valid idea.

Other factors that help towards FOTN are choke submission attempts (28 points), sweeps (26 points) and busting up your opponent's face (36 points). Did you notice that takedown attempts (successful or not) and standups from the ground aren't included? Their only value is in changing to a new position and avoiding whatever would've happened without the takedown or standup.

Finish the Fight

Finishing the fight with a knockout or submission is worth a sizeable 58 and 66 points each. But you don't want to finish too soon because you won't get to rack up enough activity points. For some perspective, only 7 percent of FOTN awards have gone to fights finished in the first round, 22 percent to fights finished in the second round, and the remaining 71 percent to fights ending in the third round or later. Getting a TKO or submission in the final round is ideal.

Title Fights and Other Awards

Title fights may have heightened anticipation and build-up making FOTN more likely. Winning Knockout of the Night or Submission of the Night - and now Performance of the Night - could hurt your chances of winning FOTN if the UFC wants bonus awards spread out or doesn't want to appear to favor one fighter with multiple awards. However, none of these appears to matter. The analysis shows that title fights and winning other awards don't have any effect on your chances of winning FOTN.

Card Placement

As already mentioned, the simple statistics seem to suggest that card placement affects your odds of winning FOTN. In reality, simple is misleading as there's no statistical difference between the main event and the rest of the main card (spots 1-5). However, fighting in one of the final three bouts of the prelims (spots 6-8) comes at a cost of 89 points. Fighting very early in the prelims (spot 9 and above) is hazardous to your FOTN health, coming at a huge cost of 141 points.

So is the UFC favoring well-known, popular fighters on the main card? It's possible. But a more likely explanation is that UFC decision makers are being influenced by things like the crowd, and maybe the recency of the fights. Crowd noise and size are very important factors that have been shown to affect the decisions of officials in other sports like the NBA and professional soccer. In the UFC, most fans don't arrive until the start of the main card. Would a fight seem more exciting to you with a packed crowd cheering and reacting to big strikes or with a puny crowd and silence?

Recency bias is a tendency to be over-affected by things that happened more recently. It plays a role in American Idol voting (better to sing last) and I've found evidence consistent with it in MMA judging (more likely to select the winner of the previous round to win the next round). The closer a fight is to the main event, the more recent the performance and therefore the more influence on whoever decides FOTN. However, this explanation doesn't fit well with the data since there's no statistical difference between main card spots 1-5. What's relatively constant throughout the main card? A large crowd making noise for exciting fights.

Here's a summary of the formula for winning Fight of the Night:

Let's revisit Fights A and B which were both from UFC 155. Fight A is Max Holloway vs. Leonard Garcia and Fight B is Jim Miller vs. Joe Lauzon.

While Holloway vs. Garcia had striking advantages and was a closer fight, the takedown advantage doesn't help towards FOTN. Meanwhile, Joe Lauzon was battered and bloody for all three rounds against Jim Miller. More chokes were attempted and, when it came time to decide FOTN, Miller and Lauzon had performed in the co-main event in front of a larger, louder crowd than Holloway and Garcia (early undercard purgatory).

Let's finish with a look at two more fights, this time from UFC 124. Matt Riddle and Sean Pierson packed the same or more activity and excitement into three rounds as Georges St-Pierre and Josh Koscheck packed into five. Riddle vs. Pierson was a much closer fight and overcame the undercard point reduction to have a 60 percent chance of winning FOTN, by far the best odds of any fight that night. GSP vs. Koscheck had only a 2.5 percent chance to win the award and ended up getting it. Why? You may remember that FOTN was decided by fan voting and boy did we screw it up - costing Riddle and Pierson \$100K each (Riddle's dispensary must have been so sad). Dana White likely would have chosen Riddle vs. Pierson for FOTN and the numbers support him.

Analyzing people's decisions - FOTN decisions in this case - isn't easy, but the advanced stats revolution in sports is allowing us to gain real insights into areas where we previously had to go with simple stats, hunches, or guesses. MMA statistics are only about six years old so, compared to other sports, we're just getting started when it comes to analyzing them. FOTN has a formula and today the analytics helped reveal what it is.

Are there any a fight questions you'd like analyzed? Send them to paul.gift@pepperdine.edu or @MMAanalytics on Twitter. Fight data provided by FightMetric.

## Trending Discussions

forgot?

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

I already have a Vox Media account!

### Verify Vox Media account

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

We'll email you a reset link.

Try another email?

### Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

### Join Bloody Elbow

You must be a member of Bloody Elbow to participate.

We have our own Community Guidelines at Bloody Elbow. You should read them.

### Join Bloody Elbow

You must be a member of Bloody Elbow to participate.

We have our own Community Guidelines at Bloody Elbow. You should read them.